The relations between a high summer temperature and excessive mortality from diarrhoea have long been well known, but the immediate cause of the disease as an epidemic is not known. Summer diarrhoea prevails to a greater extent in certain localities, notably in Leicester (and has done so for years); and the cause has been carefully sought for, but has not been found out. Recent researches, however, point to a kind of bacillus as the immediate cause, as it has been found in the air of water-closets, in the traps under the pans, and in the discharges from infants and young children. In order to indicate more readily how intimately the mortality from diarrhoea depends on temperature, I now lay before you a table showing the mean temperature for ten weeks in summer, of seven cold and hot summers, the temperature of Thames water, and the death-rates of infants under one year per million population of London: _London.--Deaths under 1 Year, in July, August, and part of September, from Diarrhoea per 1,000,000 Population Living at all Ages, arranged in the Order of Mortality._ Age 0-1 year. Mean Temperature Deaths from Diarrhoea Years. temperature, of Thames per 1,000,000 10 weeks. water. population living at all ages. 1860 58.1° 60.6° 151 1862 59.0 62.0 189 1879 58.7 60.7 228 1877 61.2 63.3 347 1874 61.7 63.8 447 1878 63.7 64.1 576 1876 64.4 64.9 643 As may be seen, the deaths of infants under 1 year of age from diarrhoea per 1,000,000 population was only 151; while the mean summer temperature was only 58.1° F. against 189 in 1862, when the mean temperature was 59.0°. In 1879, when the mean temperature was 58.7°, the deaths from diarrhoea rose to 228 per million, but a few days were unusually hot. In 1877 the mean temperature of the air was 61.2°, of the Thames water 63.3°, and the mortality of infants from diarrhoea 347 per million population. In 1874, when the mean temperature of the air was 61.7°, the mortality rose to 447 per million; and in the hot summers of 1878 and 1876, when the mean air temperatures were 64.1° and 64.9° respectively, the death-rates of infants were 576 and 642 per million population. The relations, therefore, between a high summer temperature and the mortality from diarrhoea in infants are very intimate. I have selected the mortality among infants in preference to that at all ages, as the deaths occur more quickly, and because young children suffer in greater proportion than other persons.